Mesoscale Convective Complexes and Persistent Elongated Convective Systems Over the United States During 1992 and 1993

578 MONTHLY WEATHER REVIEW VOLUME 126

Mesoscale Convective Complexes and Persistent Elongated Convective Systems over the United States during 1992 and 1993

CHRISTOPHER J. ANDERSON AND RAYMOND W. A RRITT Department of Agronomy, Iowa State University, Ames, Iowa (Manuscript received 5 February 1997, in ®nal form 12 June 1997)

ABSTRACT Large, long-lived convective systems over the United States in 1992 and 1993 have been classi®ed according to physical characteristics observed in satellite imagery as quasi-circular [mesoscale convective complex (MCC)] or elongated [persistent elongated convective system (PECS)] and cataloged. The catalog includes the time of initiation, maximum extent, termination, duration, area of the Ϫ52ЊC cloud shield at the time of maximum extent, signi®cant weather associated with each occurrence, and tracks of the Ϫ52ЊC cloud-shield centroid. Both MCC and PECS favored nocturnal development and on average lasted about 12 h. In both 1992 and 1993, PECS produced Ϫ52ЊC cloud-shield areas of greater extent and occurred more frequently compared with MCCs. The mean position of initiation for PECS in 1992 and 1993 followed a seasonal shift similar to the climatological seasonal shift for MCC occurrences but was displaced eastward of the mean position of MCC initiation in 1992 and 1993. The spatial distribution of MCC and PECS occurrences contain a period of persistent development near 40ЊN in July 1992 and July 1993 that contributed to the extreme wetness experienced in the Midwest during these two months. Both MCC and PECS initiated in environments characterized by deep, synoptic-scale ascent associated with continental-scale baroclinic waves. PECS occurrences initiated more often as vigorous waves exited the intermountain region, whereas MCCs initiated more often within a high-amplitude wave with a trough positioned over the northwestern United States and a ridge positioned over the Great Plains. The low-level jet transported moisture into the region of initiation for both MCC and PECS occurrences. The areal extent of convective initiation was limited by the orientation of low-level features for MCC occurrences.

1. Introduction vided a climatological reference for MCC occurrences. These summaries indicate that on average, about 35 Mesoscale convective complexes (MCC; Maddox MCC occurrences have affected the United States an- 1980) are large, long-lived convective systems that ex- nually, with a maximum of 59 in 1985 (Augustine and hibit a quasi-circular cloud shield (infrared temperature Howard 1988) and a minimum of 23 in 1981 (Maddox ՅϪ52ЊC) in infrared (IR) satellite imagery. Table 1 lists the speci®c criteria that must be met in order to classify et al. 1982). Typically, the southern Great Plains ex- a convective system as an MCC. These criteria differ periences most of the MCC activity during the spring, slightly from the origional criteria established by Mad- while the MCC activity shifts northward during the sum- dox (1980) in order to simplify the documentation pro- mer. During the fall, the bulk of the MCC activity returns cedure (Augustine and Howard 1988). Research has to the southern Great Plains. The distribution of MCC shown that MCCs can induce adjustments by the large- occurrences during a dry hydrological anomaly in the scale atmospheric circulation (e.g., Fritsch and Maddox central United States in 1983 did not follow the cli- 1981; Menard and Fritsch 1989) and can produce haz- matological trends. Rather, a persistent anticyclone over ardous weather over a large region. Thus, MCCs rep- the central United States hindered the moisture ¯ux into resent a challenging and important forecast problem. the Great Plains that aids the development of MCCs. Annual summaries of MCC occurrences over the This resulted in an increased frequency of smaller quasi- United States (Maddox 1980; Maddox et al. 1982; circular convective system occurrences north of the Rodgers et al. 1983; Rodgers et al. 1985; Augustine and Great Plains throughout the summer and fall (Fritsch et Howard 1988; Augustine and Howard 1991) have pro- al. 1986). These summaries have also contributed knowledge of large-scale features that support MCC development. Au- gustine and Howard (1988) used monthly mean Q-vec- Corresponding author address: Christopher J. Anderson, Depart- tor divergence (Hoskins et al. 1978) to compare the ment of Agronomy, 3010 Agronomy Building, Iowa State University, Ames, IA 50011. synoptic-scale vertical motion for an inactive warm sea- E-mail: [email protected] son (1984) with a very active warm season (1985). Their

᭧ 1998 American Meteorological Society

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 579

TABLE 1. Modi®ed MCC de®nition. ful®lls the size and duration criteria, but not the shape Size Continuous cold-cloud shield (IR temperature criterion, of the MCC de®nition. That is, a PECS must ՅϪ52ЊC) must have an area Ն50 000 km2 meet every criterion in Table 1 except the ratio of the Initiation Size de®nition is ®rst satis®ed minor to major axis does not exceed 0.7. While the Duration Size de®nition must be met for a period of internal circulation of convective systems is not ad- Ն6h Maximum extent Continuous cloud shield (IR temperature dressed with a satellite-based de®nition (as pointed out ՅϪ52ЊC) reaches maximum size by Maddox 1980), observational studies (e.g., Fritsch Shape Minor axis/major axis Ն0.7 at time of and Maddox 1981), as well as the duration and size of maximum extent these convective systems, suggest that most contain a Termination Size de®nition is no longer satis®ed distinct, mesoscale circulation. In order to diminish the likelihood that extremely long lines of convective ele- ments (convective systems not likely to contain a dis- results showed stronger synoptic-scale ascent at 600 and tinct, mesoscale circulation) meet our PECS criteria, we 400 mb for June 1985 (the month with the largest dif- established a lower limit of 0.2 for the ratio of the minor ference in MCC occurrences between 1984 and 1985) to major axis. Our comparative study consists of a tab- suggesting that this period of persistent MCC devel- ulation of PECS and MCC occurrences in 1992 and opment was supported by a persistent, upper-level 1993 and composites of environmental features near the mass±momentum imbalance in the synoptic-scale cir- time of initiation. The tabulation of MCC occurrences culation. Augustine and Howard (1991) addressed low- for 1993 has added to the climatological reference base level features supporting MCC development by com- an account of MCC occurrences during an extreme wet positing 850-mb wind, mixing ratio, geopotential hydrological anomaly in the Great Plains [for an over- height, and 500-mb geopotential height for weeks of view of the 1993 ¯ood see e.g., Kunkel et al. (1994)]. frequent MCC occurrences and weeks of infrequent MCC occurrences in 1986 and 1987. The active weeks were dominated by a deep tropospheric ridge centered 2. Data and methodology over the southeastern United States with low-level mois- a. Data sources ture ¯ux into the central plains region. Most MCC occurrences were observed near the intersection of a low- Cloud-top characteristics in GOES-7 IR digital sat- level jet (LLJ) and a zonally oriented low-level front. ellite imagery for 1992 and 1993 were analyzed with a The inactive periods were dominated by a 500-mb version of the automated routine developed by Augus- trough over the southeastern United States and low-level tine (1985). The routine searches all pixels within the moisture transport into that region. image that represent cloud-top temperature less than or Unfortunately, large, long-lived non-MCC convective equal to Ϫ52ЊC and uses a harmonic analysis technique system occurrences have not been documented as thor- to compute the best-®t ellipse for the Ϫ52ЊC cloud- oughly despite a potential for hazardous weather similar shield perimeter. The Ϫ52ЊC cloud-shield centroid and to that of MCC occurrences. Only one of the annual eccentricity are computed from the best-®t ellipse. We MCC summaries (Maddox et al. 1982) has included a also view imagery for each system to check that the list of ``non-MCC signi®cant mesoscale convective automated procedure does not detect spurious systems. events.'' The list does not allow for a comparison with We then tabulate the duration of the MCC or PECS the MCC events because satellite characteristics and a occurrence, size of the Ϫ52ЊC cloud shield at maximum description of the life cycle of each event were not extent, and hourly position of the Ϫ52ЊC cloud-shield included. Bartels et al. (1984) performed a satellite- centroid. In addition, we have summarized signi®cant based climatology for 1978±83 over the Stormscale Op- weather associated with each MCC and PECS occur- erational and Research Meteorology (STORM)-Central rence from the National Oceanic and Atmospheric Ad- project domain that included convective systems ex- ministration/National Environmental Satellite, Data, ceeding 250 km along their major axis and 3 h during and Information System (NOAA/NESDIS) publication their life cycle. MCC occurrences and large, long-lived Storm Data. linear convective system occurrences ®t within these Archived data from the National Meteorological Cen- criteria. In their tabulation, large, long-lived linear con- ter (NMC) Global Data Analysis System (GDAS; Kal- vective systems (their ``large cloud lines'') totaled about nay et al. 1990) were used to construct composite anal- one-third of the number of MCC occurrences. yses near the times of MCC and PECS initiation. In We address the lack of information concerning large, both 1992 and 1993, the GDAS used the T80 resolution long-lived non-MCC mesoscale convective system oc- (equivalant to about 200-km grid spacing) Medium currences using a comparative study with MCC occur- Range Forecast Model to provide the background or rences in 1992 and 1993. We refer to a large, long-lived ``®rst-guess'' ®elds for the analyses. Over data-rich ar- non-MCC mesoscale convective system as a ``persistent eas such as North America, the GDAS analyses mainly elongated convective system'' (PECS). Formally, we re¯ect the observations with relatively little in¯uence de®ne a PECS as a mesoscale convective system that from the model-based ®rst-guess ®elds. This dataset

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 580 MONTHLY WEATHER REVIEW VOLUME 126 contains 2.5Њ latitude±longitude grids at 0000 and 1200 UTC of surface temperature, geopotential height, and wind at 1000, 850, 700, 500, 400, 300, 250, 200, 150, 100, 70, and 50 mb. Relative humidity is also included at levels from 1000 through 300 mb. b. Fixed-point compositing methods Composites of atmospheric parameters for June±July 1993 were constructed by averaging data from 0000 UTC GDAS grids at each grid point over every day in June and July. Data were taken solely from 0000 UTC grids in order to limit the in¯uence of the diurnal evo- FIG. 1. Response function of the Barnes analysis routine applied to lution of the atmospheric circulation and the in¯uence the composite storm-relative data. of large convective systems. A deviation from the background composite on days when MCC or PECS occurrences initiated between 1800 and 0600 UTC (within the raw values. A variant of the Barnes (1994) objective 6 h of 0000 UTC) in June±July 1993 was computed by analysis routine was applied to the composited data [fol- differencing grids on these days from the background lowing the approach suggested by Achtemeier (1989)] composite grids. The deviations were then averaged in order to map the composited data onto the stereo- over the number of MCC or PECS occurrences to pro- graphic grid. The response function for our implemen- duce MCC and PECS deviation composites. tation of the Barnes scheme is shown in Fig. 1. Once Augustine and Howard (1991) described this as a the data were mapped onto the stereographic grid, the ``®xed point'' compositing technique and pointed out mean of the raw center gridpoint value was added to that this technique retains persistent large-scale features the composited deviation values in order to produce while smoothing out recurrent mesoscale features. This ®elds of meteorologically relevant magnitudes. Derived compositing technique allows for comparisons between terms, like advection, vorticity, and Q-vector diver- the large-scale circulation associated with the initiation gence, were computed after the meterological ®elds of MCC and PECS occurrences in June±July 1993 with were recreated from the deviations. the well-documented large-scale circulation of June± A stereographic grid was used rather than a latitude± July 1993. longitude grid in order to minimize spatial distortion in the composite. This distortion arises from the variation with latitude of the absolute distance for a constant lon- c. Storm-relative compositing methods gitudinal separation and has been recognized as a source A new storm-relative compositing technique was de- of error by Maddox (1983). The map factor for a ster- veloped in an attempt to reduce spatial distortion and eographic grid varies less with latitude than does the noise in the composite ®elds. Data from the 2.5Њ GDAS map factor for a latitude±longitude grid, reducing the grids were repositioned within a 5000 km ϫ 3000 km spatial distortion. stereographic grid centered on each convective system's MCC-relative and PECS-relative composites were centroid. (Grid spacing of the stereographic grid was generated by applying this technique to 38 MCC oc- 100 km.) We used 0000 UTC grids for convective sys- currences and 70 PECS occurrences from our tabulation. tems that initiated between 2200 and 0400 UTC and The composites do not contain all convective-system 1200 UTC grids for convective systems that initiated occurrences, because the archived GDAS analyses did between 1000 and 1600 UTC. These time ranges are not contain data on 15 July, 3 August, 20 September, designed so that our composites re¯ect environmental and 7 October in 1992 and 30 March, 1 May, and 8 conditions prior to initiation for a majority of the oc- July in 1993. currences in our tabulation. The deviations of the data values from the center gridpoint value were composited, 3. Catalog of MCC and PECS in 1992 and 1993 and separate composites were also constructed from the raw values for each occurrence of a convective system. a. MCC and PECS occurrences in 1992 Compositing the deviation values reduces the ambiguity In 1992, a total of 27 MCC and 77 PECS occurrences that arises from seasonal trends in the mean value of were observed over the United States from early March the atmospheric parameter being composited, better ful- through mid-November. The higher frequency of PECS ®lling the assumption of the composite-before-objective contrasts with Bartels et al. (1984) who found fewer analysis technique that the phenomenon and its envi- large cloud-line occurrences than MCC occurrences in ronment are similar from event to event. Therefore, 1978±83. The number of MCC occurrences in 1992 was these composites retain environmental spatial variability eight less than the annual average. Summaries of MCC while introducing less spatial noise than composites of and PECS occurrences that include the duration, max-

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 581

TABLE 2. MCC occurrences in 1992: TÐtornadoes; WÐhigh wind; LÐlightning damage; RÐheavy rain; FÐ¯ash ¯ooding; nKÐ (number of people) killed; nIÐ(number of people) injured. Case Initiation Duration Max. extent number Date(s) (Date/UTC) Maximum Termination (h) (km2) Signi®cant Wx 1 12 Apr 12/0200 12/0400 12/1000 8 79 788 No report 2 12 Apr 12/0600 12/1100 12/1400 8 145 076 H, W, F 3 16 Apr 16/0000 16/0200 16/0600 6 196 016 H, F, W 4 16 Apr 16/0100 16/0700 16/1100 10 245 191 No report 5 13±14 May 13/2100 13/0900 14/1700 20 338 200 H, W, L 6 14±15 May 14/2300 14/0700 15/1000 11 100 407 L 7 15±16 May 15/2100 15/2200 16/0400 7 68 382 W, T, H 8 15±16 May 15/2200 15/0400 16/0600 8 72 434 H, W 9 5 Jun 5/0200 5/0600 5/1400 12 111 898 H, T, F, W 10 5±6 Jun 5/2200 5/0800 6/1600 18 300 891 H , T, W, F, L 11 5 Jun 5/0500 5/0800 5/1300 8 94 417 H , T, W, F 12 20 Jun 20/0200 20/0500 20/0900 7 144 672 W, H , L 13 6 Jul 6/0600 6/0800 6/1400 8 104 717 H 14 7 Jul 7/0600 7/1500 7/1900 13 231 685 W, F, L 15 16 Jul 16/1000 16/1000 16/1500 5 147 983 H, W, F 16 22 Jul 22/0200 22/0600 22/1000 8 85 669 F 17 25±26 Jul 25/2300 26/0500 26/0800 9 202 248 W, H , F 18 30 Jul 30/0100 30/0600 30/0900 8 154 460 W, H , T, F 19 3 Aug 3/1300 3/1300 3/2300 10 107 053 W, H 20 4 Aug 4/0500 4/1000 4/1400 9 123 357 W, H 21 5 Aug 5/0100 5/1200 5/1500 14 111 616 H, W, F, L, 9I 22 6 Aug 6/0100 6/0200 6/0700 6 75 019 No report 23 9±10 Aug 9/1800 9/2000 10/0000 6 102 327 H, W, F 24 17±18 Aug 17/2300 18/0700 18/1100 12 149 491 W, H 25 3 Sep 3/0200 3/0900 3/1700 15 192 253 W, H 26 8 Oct 8/1600 8/2000 8/2300 7 98 580 No report 27 3±4 Nov 3/2300 4/0600 4/0600 7 133 802 T, W, H imum areal coverage of the Ϫ52ЊC cloud shield, and 0200±0300 UTC at the summer solstice.) Notice in Fig. associated signi®cant weather are provided in Table 2 3a that the peak frequency of MCC occurrence was and Table 3, respectively. observed in July. These ®ndings are in agreement with In general, PECSs in 1992 generated a larger Ϫ52ЊC previous MCC summaries that reported the highest fre- cloud shield and persisted longer than MCCs (Table 4). quency of MCC occurrences during summer nights (e.g., Despite this, MCC occurrences typically traveled a see Augustine and Howard 1991). The summertime greater distance based on the centroid positions at ini- peak was not as pronounced for the PECS occurrences, tiation, maximization, and termination. (We recognize which also had a secondary maximum in September that the centroids for the PECS occurrences do not take (Fig. 3b). into account the orientation of the PECS occurrences. Most of the early-season MCCs were observed over As a result, it is impossible to determine the entire region the southern portion of the United States. The mean affected by an individual PECS occurrence. However, position of MCC occurrence then shifted northward dur- because the centroid is the median latitude and longitude ing June (Fig. 4). These characteristics of the distri- of the Ϫ52ЊC cloud-top area, centroid tracks retain an bution are consistent with the climatological distribution average motion for the PECS and MCC occurrences.) of MCC occurrences. However, few MCCs were ob- The westward (i.e., positive) skewness in the distribu- served north of 40ЊN in July 1992, so that the northward tion of MCC occurrences and the eastward skewness in shift of the 1992 mean position of MCC initiation was the distribution of PECS occurrences suggest that most much less than the climatological mean seasonal shift of the PECS occurrences initiated, maximized, and ter- in position of MCC occurrence. The mean position of minated farther east of the MCC occurrences than in- PECS occurrence (Fig. 5) shifted farther northward dur- dicated by the mean positions. ing June and July 1992 than did the mean position of Both MCC and PECS frequently initiated in the late MCC occurrences and returned to the southern United afternoon and developed overnight (Fig. 2). PECS dis- States in mid-September. Similar to the distribution of played greater variability in the time of initiation with MCC occurrences, a period of persistent PECS occur- a higher percentage of occurrences initiating between rences was observed near 40ЊN in June. 0600 and 2000 UTC. We also found that the percentage of MCCs that initiated in the evening hours increased b. MCC and PECS occurrences in 1993 during the summer months. (Local sunset in the central From late March through mid-September of 1993, a United States is around 0000 UTC at the equinoxes and total of 28 MCCs and 44 PECSs were observed over

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 582 MONTHLY WEATHER REVIEW VOLUME 126

TABLE 3. As in Table 2 except for PECS occurrences in 1992. Asterisk denotes missing data. Case Initiation Duration Max. extent number Date(s) (Date/UTC) Maximum Termination (h) (km2) Signi®cant Wx 1 4 Mar 4/0100 4/0600 ∗ ∗ 258 434 T, H , W, F, L 2 6 Mar 6/1000 6/1200 6/2000 10 103 373 No report 3 9±10 Mar 9/1900 10/1700 11/0900 38 393 938 H, W, T 4 18 Mar 18/0300 18/1300 18/2200 19 175 519 H, W 5 6 Apr 6/0400 6/1000 6/1300 9 138 951 No report 6 10 Apr 10/0500 10/0800 10/1800 13 105 163 H. W 7 11 Apr 11/0000 11/0600 11/1300 13 572 739 H, W, F, L, T 8 13 Apr 13/0700 13/1200 13/1800 11 94 992 No report 9 17 Apr 17/0000 17/0600 17/1000 10 304 639 H, W, F, T 10 18 Apr 18/0400 18/0900 18/1600 12 306 917 F, W 11 19 Apr 19/0800 19/0800 19/1400 6 98 269 W, H , L 12 25 Apr 25/0300 25/1100 25/1600 13 290 023 W, H , F, L 13 2±3 May 2/1500 3/0100 3/0500 14 157 883 W 14 16 May 16/0600 16/0700 16/1400 8 295 670 T, H , W 15 16±17 May 16/2100 17/0000 17/0400 7 163 153 F, T, W, H 16 22±23 May 22/2000 22/2300 23/1600 20 265 485 F, H , T, W 17 31 May 31/0800 31/1000 31/1600 8 104 385 No report 18 1 Jun 1/0500 7/0700 1/1100 6 82 902 No report 19 7 Jun 7/0000 7/0200 7/0600 6 107 593 L 20 9 Jun 9/0300 9/0600 9/0900 6 94 045 H , W, T, F 21 13±14 Jun 13/2100 14/0300 14/1300 16 209 053 H, L 22 14±15 Jun 14/1000 14/1400 15/0600 20 125 178 H, F, W 23 15±16 Jun 15/2300 16/0500 16/1500 16 340 871 H, T, L, W, F 24 16±17 Jun 16/2300 17/0500 17/1400 15 242 395 H, T, W 25 17±18 Jun 17/2000 17/2300 18/1200 16 133 400 W, H, L, T, F 26 19±20 Jun 19/1700 20/000 20/0400 11 126 079 H , W, T, F, L 27 19±20 Jun 19/2300 20/0200 20/0600 7 142 431 H , W, T, F, L 28 21 Jun 21/0300 21/0300 21/0800 5 129 403 No report 29 1 Jul 1/0000 1/0400 1/1100 11 235 181 H, W, T 30 2 Jul 2/0400 2/1300 2/1700 13 164 798 H, W, L, F, T 31 3±4 Jul 3/1800 3/2200 4/0100 7 128 445 H 32 3±4 Jul 3/2000 4/0000 4/0200 6 98 638 T, L , W, H , F 33 4±5 Jul 4/2100 5/0400 5/1300 16 347 244 H, L, T, W 34 5 Jul 5/0400 5/0900 5/1800 14 131 859 L, W, H 35 8 Jul 8/0300 8/0800 8/1900 16 321 994 H, W, L 36 9 Jul 9/0100 9/0600 9/0900 8 75 957 H 37 11 Jul 11/0100 11/0300 11/0700 6 115 479 H 38 15±16 Jul 15/1800 15/1900 16/0100 7 111 949 T 39 16 Jul 16/0100 16/0400 16/0800 7 179 230 T, W, H , F, L 40 18 Jul 18/0200 18/0600 18/0900 7 103 576 F, H 41 19 Jul 19/0000 19/0500 19/0700 7 97 381 H, W, T 42 24±25 Jul 24/2300 25/0200 25/1100 12 105 707 T, F, L 43 26±27 Jul 26/2200 27/0300 27/0600 8 169 826 H, L, W 44 26±27 Jul 26/2200 27/0600 27/0600 8 190 076 L, F, W 45 30±31 Jul 30/2300 31/0200 31/0800 9 104 352 H, F, W, L 46 7 Aug 7/0000 7/0600 7/1400 14 326 962 H, W, L 47 8 Aug 8/1200 8/1600 8/1900 7 98 583 F, L 48 10 Aug 10/0200 10/0800 10/1300 11 175 736 T 49 10±11 Aug 10/2200 11/0700 11/1200 14 286 851 H, W, F 50 11±12 Aug 11/2300 12/000 12/0800 9 103 750 L, F, H, W 51 2 Sep 2/0000 2/0400 2/0900 9 83 057 No report 52 5 Sep 5/0500 5/0900 5/1200 7 194 880 No report 53 5 Sep 5/1100 5/1100 5/1800 7 118 575 W, H 54 6 Sep 6/0200 6/0500 6/1200 10 280 833 T, H , F, W 55 7 Sep 7/0300 7/0600 7/1100 8 131 929 H, L 56 7 Sep 7/0300 7/0900 7/1800 15 152 010 T, H , L 57 8 Sep 8/0300 8/0500 8/0900 6 250 321 W, L 58 8±9 Sep 8/1800 9/0100 9/0500 11 79 773 F, H 59 9±10 Sep 9/0900 10/0300 10/0900 24 423 077 H, W, L, F 60 10±11 Sep 10/1100 11/0100 11/1100 24 209 775 H, W, L, F 61 14 Sep 14/0600 14/1200 14/1900 13 388 828 F 62 15 Sep 15/0200 15/0600 15/1500 13 132 215 F 63 16 Sep 16/0100 16/1100 16/1700 16 258 180 H, W, F, L 64 17 Sep 17/0600 17/1000 17/1200 7 110 346 W 65 18 Sep 18/0200 18/0900 18/1500 13 195 840 H, W 66 19 Sep 19/0900 19/1400 19/1800 9 148 982 No report 67 20 Sep 20/0200 20/0400 20/1000 8 112 068 W, H , L

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 583

TABLE 3. (Continued) Case Initiation Duration Max. extent number Date(s) (Date/UTC) Maximum Termination (h) (km2) Signi®cant Wx 68 7 Oct 7/0500 7/0900 7/1100 6 158 516 R, F 69 9 Oct 9/0200 9/0800 9/1300 11 144 491 W, H , L 70 9 Oct 9/0200 9/1000 9/1200 10 159 159 R 71 30 Oct 30/0000 30/0500 30/1300 13 237 027 H, W 72 1 Nov 1/0500 1/0700 1/1300 8 194 383 W, H , T, L 73 1±2 Nov 1/1100 2/0200 2/0500 18 153 558 W, T, H 74 2±3 Nov 2/0400 2/1300 3/1300 33 270 372 W 75 11 Nov 11/0600 11/1200 11/1500 9 76 414 No report 76 12±13 Nov 12/0800 13/0400 13/1500 31 297 526 F, W 77 20±22 Nov 20/2100 21/0600 22/1200 39 348 530 T, W, H , F, 2 6 K, 500ϩ I the United States. Similar to 1992, MCC occurrences quency of occurrence was observed in July (Fig. 3a). were less frequent than the annual average (by seven), In addition, a larger percentage of MCCs in the summer and PECSs were more frequent than MCCs. Tables 5 months initiated in the late afternoon than those in the and 6 provide summaries for MCC and PECS occur- spring months. The month of peak frequency for PECS rences, respectively, that include the duration, maximum occurrences was June. Unlike the MCC occurrences, the areal coverage of the Ϫ52ЊC cloud shield, and associ- percentage of PECS occurrences initiating in the eve- ated signi®cant weather. ning did not change much between spring and summer. The mean PECS occurrence was larger than the mean PECSs were more frequent during the spring and fall MCC, while MCCs in general were longer lived and than MCCs. traversed a greater distance than PECSs (Table 4). PECS Similar to 1992, PECS occurrences in 1993 (Fig. 7) occurrences initiated, maximized, and terminated farther followed a more pronounced seasonal migration of the east and south than MCCs. Both MCCs and PECS fa- mean position than did the MCCs (Fig. 6). Both distri- vored nocturnal development (Fig. 2). The peak fre- butions showed a concentration of systems near 40ЊN. This

TABLE 4. Selected MCC and PECS occurrence statistics [skewness] {kurtosis}. 1992 1993 MCC PECS MCC PECS Number of occurrences 27 77 28 44 Maximum extent of 145 098 188 009 184 285 193 167 Ϫ52ЊC cloud shield [1.259] [1.290] [1.171] [0.705] (km2) {0.939} {1.665} {1.522} {Ϫ0.337} Duration (h) 9.6 12.2 13.9 12.4 [1.273] [2.047] [1.859] [1.056] {1.015} {4.584} {3.729} {0.981} Initiation position (north 38.81, 100.47 39.09, 95.17 40.80, 100.76 38.73, 97.01 lat, west long) [0.221, 0.442] [Ϫ0.053, Ϫ0.292] [Ϫ0.938, Ϫ0.064] [0.070, Ϫ0.606] {Ϫ0.393, Ϫ0.146} {Ϫ1.010, 0.096} {0.373, Ϫ0.137} {Ϫ0.600, 0.180} Maximum position (north 38.28, 98.15 38.94, 93.39 40.84, 96.94 38.83, 94.52 lat, west long) [0.164, 0.494] [Ϫ0.179, Ϫ0.418] [Ϫ0.754, 0.458] [0.009, Ϫ0.279] {Ϫ0.100, Ϫ0.314} {Ϫ0.960, 0.083} {Ϫ0.085, Ϫ0.210} {Ϫ0.377, 0.259} Termination position 37.96, 95.61 39.37, 90.44 40.04, 93.19 38.61, 90.89 (north lat, west long) [Ϫ0.018, 0.258] [Ϫ0.004, Ϫ0.482] [Ϫ0.918, Ϫ0.336] [Ϫ0.051, Ϫ0.195] {Ϫ0.409, Ϫ0.081} {Ϫ0.768, 0.218} {0.545, 1.100} {0.347, 0.240} All MCC All PECS Initiation position (north 39.83, 100.62 38.95, 95.84 lat, west long) [Ϫ0.335, 0.249] [Ϫ0.008, Ϫ0.414] {Ϫ0.460, Ϫ0.056} {Ϫ1.158, Ϫ0.444} Maximum position (north 39.59, 97.53 38.90, 93.80 lat, west long) [Ϫ0.287, 0.593] [Ϫ0.124, Ϫ0.447] {Ϫ0.460, 0.076} {Ϫ0.823, Ϫ1.375} Termination position 39.20, 94.38 39.09, 90.60 (north lat, west long) [Ϫ0.444, 0.043] [0.029, Ϫ0.482] {Ϫ0.265, 0.624} {Ϫ0.794, Ϫ1.136}

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 584 MONTHLY WEATHER REVIEW VOLUME 126

FIG. 2. Hourly distribution of the time of initiation (UTC) for (a) MCC and (b) PECS occurrences in 1992 (solid) and 1993 (hatched).

FIG. 3. Monthly distribution of (a) MCC and (b) PECS occurrences spatial concentration of MCC occurrences (re¯ected in for 1992 (solid) and 1993 (hatched). Table 4 as positive kurtosis) was more pronounced, extending through much of the summer, whereas the persistence of PECS occurrences spanned the ®rst half of June. than MCC occurrences. The mean centroid positions do not fully describe this difference as the distribution of PECS occurrences contains a bias eastward of the mean c. General comparison of MCC and PECS centroid positions. In fact, most MCC occurrences ter- occurrences minated prior to 95ЊW (about the mean initiation position Bartels et al. (1984) found that large cloud lines and of PECS occurrences) as evidenced in the distribution of MCCs in their climatology displayed similar diurnal char- MCC occurrences by the westward skewness and clus- acteristics. Our results support their ®nding as both types tering about the mean longitude of termination (Table 4). of systems favored initiation in the late afternoon (Fig. 2) MCC and PECS occurrences in 1993 on average lasted and developed overnight, typically lasting about 12 h (Ta- longer and were larger than in 1992 (Table 4). Cotton et ble 4). Bartels et al. (1984) reported that the occurrence al. (1989) described the MCC as an inertially stable con- of large cloud lines increased during the summer, similar vective system, or a system whose spatial scale is com- to our ®ndings, but their summertime peak was not as parable to, or exceeds, the Rossby radius of deformation pronounced as the peak in our tabulation. In addition, our causing the system to approach geostrophic balance. This combined tabulation for 1992 and 1993 reveals a second description is supported by numerical simulation (Zhang peak in the frequency of PECS occurrence in September and Fritsch 1987) and observational evidence (Menard and (Fig. 3). Their study was limited to convective system Fritsch 1989). The correlation between size and duration occurrences prior to August over the STORM-Central do- of MCC occurrences found in this and other studies (e.g., main. Laing and Fritsch 1997) agrees with this description, be- PECS occurrences initiated and developed farther east cause the likelihood that the spatial scale of the MCC

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 585

FIG. 4. Tracks of Ϫ52ЊC cloud-shield centroids for MCC occurrences in 1992 during (a) March, April, May; (b) June, July, August; and (c) September, October, November.

exceeds the Rossby radius of deformation increases with 1992 as one of seven hydrometeorological factors that the size of the MCC (all else being equal). The corre- contributed to the regional ¯ooding observed in the spondence between the mean PECS size and duration in Midwest during the summer of 1993. Our tabulation our tabulation suggests the possibility that inertially stable contains an extraordinary number of PECS occurrences vortices may play a role in the persistence of PECS oc- in June±July 1992 that nearly equaled the total number currences as well as MCCs. However, the mean MCC of large cloud lines in June 1979±83 (13 compared with occurrence in 1993 persisted longer than the larger mean 14) and surpassed the total number of large cloud lines PECS occurrences in 1992 and 1993, suggesting that this in July 1978±83 [16 compared with 12; Bartels et al. mechanism may not be as effective for PECSs as for (1984)]. Most of the PECSs in late June±July 1992 MCCs. Further study of the internal circulations within tracked over the eastern portions of the Great Plains and MCCs and PECSs would be needed to address this issue. aided soil saturation by contributing heavy rainfall amounts and by reducing evapotranspiration through increased cloud coverage. The anomolous late-season oc- d. Contribution to extreme wet periods in 1992 and currence of large convective systems in 1992 helped to 1993 maintain near-saturation soil wetness through the fall of Kunkel et al. (1994) cited soil wetness approaching 1992 into the spring of 1993. saturation over much of the Midwest beginning in July Another factor contributing to the 1993 Midwest ¯ood

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 586 MONTHLY WEATHER REVIEW VOLUME 126

FIG. 5. As in Fig. 4 except for PECS occurrences in 1992.

discussed in Kunkel et al. (1994) is the frequent occurrence parison of quasi-circular convective systems from 1982 of large-sized areas of moderate to heavy rain. They sub- and 1983, Fritsch et al. (1986) concluded that storm jectively determined that 70 large-sized areas of moderate precipitation frequency and the anomalous persistence to heavy rain affected the Midwest in the summer of 1993 of occurrences north of the Midwest contributed more and commented that ``many of the large rain areas were to the drought conditions over the Midwest in 1983 than of appropriate scale and shape to suggest the presence of frequency of occurrence. This notion has some addi- MCCs and MCSs.'' Our tabulation reports 43 MCC and tional support, as regional ¯ooding was not observed PECS occurrences over the Midwest during the summer in 1985. In that year, more MCC occurrences (59) were of 1993 supporting their characterization of the large-sized observed than in any other year for which the MCC rain areas. (This suggests that perhaps surface rainfall ob- climatologies have been performed; however, the MCCs servations could be used to infer a more comprehensive were less concentrated spatially than in 1983 and 1993 climatology of large long-lived convective system occur- (Augustine and Howard 1988). If considering only the rences prior to 1978.) MCC occurrences for 1993, our results would also imply As stated in the introduction, spatial persistence was that concentrated occurrences and storm precipitation demonstrated in the distribution of MCC occurrences in ef®ciency were more important factors contributing to 1983, a drought year (Rodgers et al. 1985). In a com- regional ¯ooding than frequency, because MCC occur-

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 587

TABLE 5. As in Table 2 except for MCC occurrences in 1993. Case Initiation Duration Max. extent number Date(s) (Date/UTC) Maximum Termination (h) (km2) Signi®cant Wx 1 7±9 Apr 7/0900 8/0200 9/0400 43 473 895 H , T, F, W, L 2 12±13 Apr 12/2000 12/2200 13/0200 6 69 718 H, L, F 3 13±14 Apr 13/1600 13/1900 14/0100 9 99 868 H, F 4 29 Apr 29/0000 29/0600 29/1200 12 293 510 H , T, W, L 5 4±5 May 4/2300 5/0500 5/1200 13 257 481 H, W 6 5±6 May 5/0400 5/1600 6/0700 27 177 215 W, T, L 7 5 May 5/1700 5/2000 5/2300 6 63 163 W, T, L 8 27±28 May 27/2200 28/0000 28/0400 6 121 907 H, W, L 9 28 May 28/0500 28/0800 28/1800 13 156 049 H, L, W 10 2 Jun 2/0300 2/0500 2/1000 7 70 622 H, W 11 9±10 Jun 9/1600 10/0700 10/1400 22 231 572 H, F, W 12 14 Jun 14/1200 14/1500 14/1800 6 96 161 H, L, F 13 15±16 Jun 15/1600 15/1900 16/0100 9 84 873 H, L, T 14 15±16 Jun 15/2300 16/0700 16/1800 19 351 442 H, W, L, F, T 15 23±24 Jun 23/2200 24/0600 24/1300 15 229 260 W, T, F, H 16 29±30 Jun 29/0200 30/0700 30/0700 29 207 246 W, H, L, T, F 17 1 Jul 1/0000 1/0600 1/1100 11 313 246 F, H, L, T 18 3 Jul 3/0400 3/0600 3/2200 18 151 900 T, H , W, F, L 19 6 Jul 6/1200 6/1800 6/2100 9 107 767 W, F, H , T 20 7 Jul 7/0400 7/0900 7/1400 10 165 263 T, F, W 21 8 Jul 8/0400 8/0700 8/1200 8 153 173 F, W, H, L, T 22 21±22 Jul 21/2300 22/0300 22/1000 11 146 776 H , T, W, F 23 23 Jul 23/0500 23/1400 23/1500 10 152 263 F 24 24 Jul 24/0200 24/1100 24/1400 12 193 912 H, F 25 26±27 Jul 26/2300 27/0400 27/1100 12 204 262 H, T, W 26 12 Aug 12/0400 12/1000 12/1600 12 261 893 F 27 14 Aug 14/0100 14/1400 14/1800 17 152 513 F, L 28 13±14 Sep 13/0600 13/1300 14/0000 18 171 030 H, W, T

rences were slightly less frequent in June±July 1993 differential vorticity advection. Further, Bell and Jan- than the climatological mean for June±July (18 com- owiak (1995) noted that the right entrance region of a pared with 20) and slightly larger (as indicated by the mean 200-mb jet streak coincided with the region of Ϫ52ЊC cloud area) in 1993 than 1992. However, when differential positive vorticity advection. The superposi- taking into account all large, long-lived convective sys- tion of the direct thermal circulation of the jet streak tem occurrences in 1993, the frequency of occurrence (Murray and Daniels 1952; Chen and Kpaeyah 1993) and cannot be downplayed so easily. the divergence center associated with the continental scale baroclinic wave suggests the potential for anoma- 4. Analysis of environments associated with MCC lously strong synoptic-scale ascent over the Midwest. As and PECS initiation we stated in the introduction, Augustine and Howard (1988) found that the large-scale circulation favorable for a. Fixed-point composites persistent MCC development was characterized by deep, The most active 2-month period in March 1992±Sep- phased synoptic-scale ascent. The results of our tabula- tember 1993 was June±July 1993, when 24 PECSs and tion indicate that an upper-level circulation characterized 18 MCCs were observed. The mean 500-mb geopotential by synoptic-scale ascent supports frequent PECS devel- height ®eld for June±July 1993 contains a trough in the opment also. northwestern United States with a broad ridge extending Our plots of deviation 500-mb geopotential height over the remaining two-thirds of the nation (Fig. 8). Bell further support the notion that the upper-air circulation and Janowiak (1995) compared the mean 500-mb geo- associated with MCC and PECS occurrences in June± potential heights for June±July 1993 with mean June± July 1993 was characterized by synoptic-scale ascent, July 500-mb geopotential heights de®ned from a base although the orientation of the upper-air features dif- period of 1979±88. Their deviation plots reveal negative fered. Negative height anomalies in the northeastern and height anomalies as large as 60 m in the northwestern northwestern United States and a positive height anom- United States, and they suggest that the trough was part aly over the central United States signify a higher-am- of a deep tropospheric baroclinic wave. According to plitude wave pattern relative to the mean 500-mb geo- baroclinic wave theory (Holton 1992), a divergent center potential height for days in June±July 1993 when MCC is expected downwind of an upper-level trough as a result occurrences were observed (Fig. 8a). In other words, of the broad region of ascent associated with positive our deviation 500-mb geopotential height composite is

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 588 MONTHLY WEATHER REVIEW VOLUME 126

TABLE 6. As in Table 2 except for PECS occurrences in 1993. Case Initiation Max. extent number Date(s) (Date/UTC) Maximum Termination Duration (h) (km2) Signi®cant Wx 1 24 Mar 24/0100 24/0400 24/1100 10 154 516 F 2 29±30 Mar 29/2000 30/0400 30/1200 16 221 053 H, W, L, T, F 3 30±31 Mar 30/1800 30/2000 31/0000 6 174 486 H , W, T, F 4 30±31 Mar 30/2100 31/0400 31/2200 25 211 169 H, W, L 5 3±4 Apr 3/2300 4/1000 4/1400 15 317 306 H, W, L 6 19 Apr 19/0100 19/1100 19/1500 14 203 333 H, W, L 7 1 May 1/0000 1/0400 1/1100 11 243 388 W, H 8 1 May 1/1900 1/2100 2/0100 6 104 633 W, H 9 2±3 May 2/0000 2/0500 3/0500 29 232 711 H, W, T 10 4 May 4/0700 4/0800 4/1500 8 163 078 No report 11 6 May 6/0000 6/0300 6/1400 14 269 960 W, T, L 12 15±16 May 15/2300 16/0500 16/0700 8 124 294 H 13 17±18 May 17/2300 18/0400 18/1200 13 363 143 W, H , T, F, L 14 18±19 May 18/1600 18/2300 19/0400 12 335 537 W, H , T, L , F 15 22±23 May 22/2300 23/0300 23/0900 10 132 719 T, H , W, F 16 30 May 30/0100 30/0600 30/0900 8 103 964 H, W, F 17 2±3 Jun 2/2200 3/0400 3/1700 19 169 099 T, H , W 18 4±5 Jun 4/0500 4/1500 5/0000 19 329 599 H , W, T, L , F 19 5 Jun 5/0000 5/0400 5/1000 10 148 262 W, T, H 20 8 Jun 8/0000 8/0200 8/1200 12 350 217 T, W, H , L 21 8±9 Jun 8/1300 8/2000 9/0700 18 159 202 H, W, F, T 22 13 Jun 13/0600 13/1000 13/1700 11 227 482 H, W 23 14 Jun 14/0000 14/0300 14/1000 10 208 996 T, W, F, L 24 14±15 Jun 14/1800 15/0000 15/0800 14 161 656 W, H , F, L 25 19 Jun 19/0100 19/0300 19/0800 7 113 891 H, F, W 26 23 Jun 23/0000 23/0300 23/2300 23 150 179 W, H , T 27 25 Jun 25/0000 25/0200 25/1500 15 222 533 W, H , F 28 2 Jul 2/0200 2/0600 2/1000 8 208 929 F, H , T, W 29 3 Jul 3/1300 3/1400 3/2000 7 93 506 F, L , T 30 7 Jul 7/0000 7/0400 7/0800 8 226 105 H, L, W, F 31 9 Jul 9/0200 9/0500 9/0800 6 223 706 H , W, T, F 32 13 Jul 13/0100 13/1000 13/1400 13 126 866 F, L 33 21 Jul 21/0100 21/0700 21/1200 11 138 214 T, W, H , F 34 23 Jul 23/0500 23/0900 23/1100 6 114 976 F 35 24 Jul 24/0500 24/0800 24/1200 7 284 222 F 36 26±27 Jul 26/2000 27/0100 27/0400 8 208 661 W, F, L 37 6 Aug 6/0000 6/0100 6/0700 7 117 877 W, L, H, T 38 9 Aug 9/0300 9/0900 9/2100 18 110 418 T, W 39 10 Aug 10/0100 10/0800 10/1500 14 149 580 H, W, F 40 11 Aug 11/0100 11/0300 11/1500 14 122 874 H, W 41 12 Aug 12/2300 13/0300 12/1200 13 256 683 F 42 13 Sep 13/0900 13/1200 13/1700 8 118 422 L, W, F, T 43 13±14 Sep 13/2100 13/2200 14/1200 15 174 380 F, T, W 44 14 Sep 14/0100 14/0500 14/1900 18 227 487 W, T, F similar to the 500-mb geopotential height composite in uated over the Midwest on days when MCC occurrences Bell and Janowiak (1995), except that the negative were observed in June±July 1993 (Fig. 9a). Unlike the anomaly over the northwestern United States in our mean June±July 1993 200-mb wind, the upper-level cir- plots implies a larger deviation from the 1979±88 mean. culation does not contain a jet streak. Instead, assuming Augustine and Howard (1991) diagnosed a high-am- the thermal wind relation, the zonal jet stream implies plitude wave pattern in the 500-mb geopotential height a zonal low-level thermal boundary. This is consistent for periods of active MCC development in 1986 and with the large-scale composites performed by Augustine 1987 that was dominated by ridging in the central United and Howard (1991), who observed that low-level ther- States, in contrast with the dominant troughing along mal forcing and convective instability when phased ver- the United States west coast in June±July 1993. PECS tically with a mid- to upper-level anticyclone was the occurrences in June±July 1993 favored initiation down- most favorable pattern for MCC development. However, wind of a trough within a continental-scale wave as the a 200-mb jet streak is resolved in the 200-mb deviation trough exited the intermountain region. This is indicated wind for the PECS occurrences in June±July 1993 (Fig. by a broad 500-mb geopotential height minimum east 9b), suggesting that upper-level jet streak circulations of the background state trough (Fig. 8b). directly in¯uence PECS initiation more frequently than The southern edge of the 200-mb jet stream was sit- MCC initiation.

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 589

FIG. 6. As in Fig. 4 except for MCC occurrences in 1993.

b. Storm-relative composites passing the trough axis and impinging on the composite MCC position of initiation from the southwest. 1) UPPER-LEVEL FEATURES A short wavelength feature was resolved by Maddox Both MCC-relative and PECS-relative composites of (1983) upstream of his composite MCC position of ini- 200-mb relative vorticity (Fig. 10) reveal that a conti- tiation, but a storm-relative composite by Cotton et al. nental-scale wave dominates the upper-level environ- (1989) did not resolve this feature. The composite by ment prior to initiation of either type of convective sys- Cotton et al. (1989) contained more samples, but the tem. Consistent with the ®xed-point composites, the samples were limited to MCC occurrences during the continental-scale wave in the PECS-relative composite summer. Therefore, their composite contained less vari- is situated east of the continental-scale wave in the ability among samples than both Maddox (1983) and MCC-relative composite. The continental-scale wave in our composite. While the results of Cotton et al. (1989) the PECS-relative composite is also a deep tropospheric are less likely to contain nonphysical short-wavelength wave, whose westward tilt with height (Fig. 10b) re- features introduced by including a seasonal trend in the sembles the westward tilt with height expected for a composite, their results may not re¯ect the variety of developing baroclinic wave (Holton 1992). In contrast, atmospheric circulations that support MCC occurrences. the MCC-relative composite of 700-mb relative vorticity In order to examine the in¯uence of sample selection indicates the presence of a short-wavelength disturbance in our composite, we constructed two MCC-relative

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 590 MONTHLY WEATHER REVIEW VOLUME 126

FIG. 7. As in Fig. 4 except for PECS occurrences in 1993.

composites for 200- and 700-mb relative vorticity and ence of short-wavelength disturbances in the develop- 200-mb wind (not shown here). One composite con- ment of MCC occurrences. First, Velasco and Fritsch tained only MCC occurrences during the summer, while (1987) proposed that the ascending motion associated the other composite contained only MCC occurrences with short-wavelength disturbances directly enhances during spring and fall. Both composites of 200-mb rel- MCC development as suggested by the composites pre- ative vorticity exhibited the upper-level continental- sented in Maddox (1983). The second school of thought scale wave previously discussed, but the short-wave- proposed that short-wavelength disturbances promote length feature produced a much stronger signal in the low-level cold-air advection enhancing low-level fron- spring±fall MCC composite. These results suggest that togenetic circulations. The cold-air advection mecha- the discrepancy between Cotton et al. (1989) and Mad- nism does not require that short-wavelength distur- dox (1983) may result simply from the differences in bances be in close proximity to the MCC occurrence in the sample selection. either space or time to in¯uence MCC development. The latter theory is more plausible when MCC occurrences develop in rapid succession (Schwartz et al. 2) ROLE OF SHORT-WAVELENGTH DISTURBANCES 1990) and establishes consistency between the results The discussion by Augustine and Howard (1991) of Maddox (1983) and Cotton et al. (1989). summarized two main schools of thought on the in¯u- Our composites also support the notion that a short-

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 591

FIG. 8. Mean June±July 1993 500-mb geopotential height (m, thin FIG. 9. The 200-mb composite wind vectors and speed (m sϪ1) for lines) and deviation (thick lines) for days in June±July 1993, when days in June±July 1993, when (a) MCC occurrences and (b) PECS (a) MCC occurrences and (b) PECS occurrences were observed. occurrences were observed. wavelength disturbance frequently in¯uences the de- of the continental-scale wave also enhances the upper- velopment of MCCs. Upper-level divergence (Fig. 11a) level divergence associated with the continental-scale extends from near the position of the short-wavelength wave. As previously discussed, high-amplitude conti- disturbance through the in¯ection point of the continental-scale waves are often observed with frequent nental-scale baroclinic wave with a maximum directly MCC development. east of the short-wavelength feature. The short-wavelength feature enhances and expands the region of up- 3) UPPER-LEVEL DYNAMIC CIRCULATION per-level divergence within the continental-scale wave. In this way, mid- to upper-level ascent associated with Consistent with previous MCC studies, we ®nd that the short-wavelength disturbance enhances MCC initi- a deep tropospheric dynamic circulation is important to ation directly, similar to the role proposed by Velasco the development of MCCs and appears to be important and Fritsch (1987). However, short-wavelength distur- to PECSs as well (Fig. 11b). The stronger upper-level bances may indirectly in¯uence MCC development as divergence maximum in the PECS-relative 200-mb di- the short-wavelength disturbance approaches the axis of vergence (Fig. 11) suggests stronger upper-level dynam- the continental-scale trough. Vorticity advection ahead ics are associated with the systems that support PECS of the short-wavelength disturbance intensi®es the cy- development. Additional upper-level support is provid- clone and increases the amplitude of the continental- ed by the cross-jet circulation at the right entrance region scale wave. (We infer this from the de®nition of geo- of the 200-mb jet streak in both MCC-relative and strophic relative vorticity in isobaric coordinates, which PECS-relative 200-mb wind composites (Fig. 12). Re- shows that local maxima of geostrophic relative vortic- call that the ®xed-point MCC composite did not resolve ity are proportional to local minima of geopotential a 200-mb jet streak. This discrepancy may result from height. Therefore, positive geostrophic relative vorticity the smoothing that occurs with the ®xed-point method, advection generally corresponds to the movement of or it may indicate that a jet streak was not observed geopotential height minima.) Increasing the amplitude very often within the sample used for the ®xed-point

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 592 MONTHLY WEATHER REVIEW VOLUME 126

FIG. 10. The 200-mb relative vorticity (1 ϫ 10Ϫ6 sϪ1, contoured) and 700-mb relative vorticity (1 ϫ 10Ϫ6 sϪ1, shaded with interval of 2 starting with 2) for (a) MCC-relative and (b) PECS-relative composites. Asterisk denotes position of initiation for the composite convective system. composite (containing only MCC occurrences in June± when compared to our summer MCC-relative composite July 1993). The latter explanation is supported by the of 200-mb wind. fact that a stronger and more compact jet streak was We provide a measure of the synoptic-scale ascent resolved in our spring±fall MCC-relative composite with composites of Q-vector divergence at 300 and 700

→

FIG. 11. The 200-mb divergence (1 ϫ 10Ϫ6 sϪ1, contoured) and 850-mb convergence (1 ϫ 10Ϫ6 sϪ1, shaded with interval of Ϫ2 starting with Ϫ4) for (a) MCC-relative and (b) PECS-relative composites.

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 593

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 594 MONTHLY WEATHER REVIEW VOLUME 126

gustine and Caracena 1994). Many circulations that produce local maxima in meso- to synoptic-scale ascent and support convective initiation are induced by this intersection. A few examples mentioned in the literature are frontogenetic circulations (Augustine and Caracena 1994), frontal overrunning (isentropic lift), and convergence (Cotton et al. 1989). In 1993, convergence associated with the LLJ played an important role in the extremely active period in late June and early July. Arritt et al. (1997) derived a composite of the 1200 UTC 850- mb ageostrophic wind component for days when a strong LLJ (Ͼ20 m sϪ1) was observed over the central United States (29±30 June, 3±9 July, and 14 July 1993). Convergence is evident in their composite over the same region (near 40ЊN) where 11 large, long-lived convective systems were observed during the period extending 29 June±10 July. In addition to inducing ascent, the LLJ supplies mois-

ture to the MCC by advecting higher-␪e air into the region of initiation (Cotton et al. 1989). Figure 14 por- trays this scenario for the MCC-relative composite and many similar features for the PECS-relative composite. The LLJ is observed south of the composite MCC and composite PECS position of initiation. In both composites, the LLJ advects moisture from the Gulf of Mex- ico into the region of initation and eastward. The eastward extension of moisture advection highlights the importance of low-level moisture transport to the suste- nance of both types of large, long-lived convective systems. Drying is observed southwest of the initation position in both composites. The presence of westerly ¯ow and drying along, and west of, a north±south-oriented con¯uent zone that is better de®ned in the PECS composite indicates that drying in the PECS composite is associated with an airmass boundary such as a dryline or a cold front. Drying ahead of the con¯uence zone in the MCC composites indicates moisture transport from FIG. 12. (a) MCC-relative and (b) PECS-relative composites of a region southwest of the initiation position to the region 200-mb wind vector and wind speed (m sϪ1). of initiation and eastward. In our composites, PECS occurrences tended to tap a moisture source to the south- east or south of the region of initation, whereas MCC mb (Fig. 13). A region of deep synoptic-scale ascent occurrences tapped moisture south or southwest of the that tilts slightly north and west with height is associated region of initiation. Moisture advection in the MCC with the deep tropospheric dynamic circulation in both composite is bounded to the north by a region of con- MCC and PECS composites. Deep synoptic-scale ascent ¯uence that probably represents a warm or stationary has been diagnosed near the time of initiation of MCC front. This boundary is more de®ned in the MCC com- occurrences by Maddox (1983) and Cotton et al. (1989) posite and con®nes moisture transport over the region. and is consistent with the ®xed-point composites we This east±west boundary may also limit the areal ex- presented earlier. However, in the PECS composite, the tent of convective initiation. The boundary is situated synoptic-scale ascent is somewhat stronger, supporting within the region of the upper-level synoptic-scale as- our earlier assertion that, in general, dynamic forcing is cent (Figs. 13a and 14a). Because conditional instability stronger with PECS occurrences than MCC occurrences. tends to be less likely on the cold side of low-level thermal boundaries, the boundary limits the northward extent over which upper-level ascent aids convection. 4) LOW-LEVEL FEATURES In comparison, the north±south boundary in the PECS- MCCs often initiate near the intersection of an LLJ relative composite does not limit the northward or south- (low-level jet) and a zonally oriented quasi-stationary ward extent over which upper-level ascent may aid con- front or warm front (Augustine and Howard 1991; Au- vection.

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 595

FIG. 13. The 300-mb Q-vector divergence (1 ϫ 10Ϫ16 sϪ3 mbϪ1, contoured) and 700-mb Q-vector divergence (1 ϫ 10Ϫ16 sϪ3 mbϪ1, shaded with interval of Ϫ1 starting with Ϫ1) for (a) MCC-relative and (b) PECS-relative composites.

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 596 MONTHLY WEATHER REVIEW VOLUME 126

FIG. 14. Streamline analysis derived from the recononstructed 850-mb wind and 850-mb moisture advection [1 ϫ 10Ϫ2 gkgϪ1 hϪ1, shaded with intervals of ՅϪ6 (lightest), Ϫ6toϪ3, Ϫ3 to 3 (no shading), 3to6,Ն6 (darkest)] for (a) MCC-relative and (b) PECS-relative composites.

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 597

Convection may also be hindered by the presence of streamline composite extends a greater distance south low-level inversions. Carlson and Farrell (1983) intro- of the initiation position than in the PECS composite. duced a lid-strength term using the difference between Therefore, acceleration of the low-level branch of the the maximum saturated wet-bulb potential temperature divergent circulation in the MCC-relative composite within the inversion and the surface wet-bulb potential may re¯ect additional physical processes. Stull (1988) temperature (term B of the lid-strength index). Because describes many boundary layer processes that induce our dataset consists solely of mandatory-level data, we low-level wind maxima. For example, differential heat- used the 700-mb saturated wet-bulb potential temper- ing over sloped terrain induces a surface pressure gra- ature in place of the maximum saturated wet-bulb po- dient and accelerates the low-level ¯ow. Consistent with tential temperature within the inversion and replaced the this boundary layer mechanism, the LLJ is positioned surface wet-bulb potential temperature with the 850-mb over sloped terrain within a moisture gradient (Fig. 14a). wet-bulb potential temperature. While our substitutions Or, the LLJ may be related to other synoptic-scale cir- prohibit a direct comparison with lid-strength threshold culations such as the westward extension of the Ber- values for surface-based convection (Farrell and Carlson muda high, which is often observed with MCC occur- 1989), the general result that convective inhibition in- rences (Augustine and Howard 1991). A more complete creases with the lid-strength term is retained. discussion of boundary layer and synoptic-scale in¯u- The initiation position in the MCC composite lies ences on the LLJ is provided in Augustine and Caracena immediately eastward and northward of a region of in- (1994) and Stensrud (1996). creasing lid strength (Fig. 15a), resembling the underrunning process associated with some severe convective 5. Summary storms (Farrell and Carlson 1989). This region of increasing lid strength hinders convection and limits the We have cataloged 27 MCC occurrences and 77 PECS southward extent of the convective system. In contrast, occurrences in 1992, and 28 MCC occurrences and 44 a local minimum and weaker gradient of the lid strength PECS occurrences in 1993. Both MCC and PECS oc- is observed near the initiation position in the PECS currences tended to be nocturnal and were observed composite. This implies that convection is not as effec- most often during the summer. The average duration tively inhibited, especially if the meridionally oriented and maximum areal coverage of the Ϫ52ЊC cloud shield con¯uence is interpreted as a thermal boundary. of MCC and PECS occurrences were comparable. The In both the MCC and PECS composites, the LLJ ap- spatial distribution of PECS occurrences displayed a pears to be induced in part by the upper-level dynamic mean seasonal shift similar to the climatological distri- forcing (Uccellini and Johnson 1979; Uccellini 1980; bution of MCC occurrences but displayed greater vari- Arritt et al. 1997). Figure 15 reveals a southerly com- ability and a tendency for PECS to initiate and develop ponent in the deviation wind ®eld that is in close prox- farther east than MCCs. PECS occurrences displayed imity to the vertically phased divergence and conver- more variability in time of initiation, position of initi- gence in Fig. 11. In a composite of the springtime LLJ ation, size, and duration than MCC occurrences. Spatial (Chen and Kpaeyah 1993), the composite LLJ was po- coherence in the distribution of both MCC and PECS sitioned within the low-level branch of the divergent occurrences, and the frequency of PECS occurrences, circulation associated with a developing continental- contributed to the extreme wet period extending from scale baroclinic wave. The exit region of an upper-level July 1992 to August 1993 that culminated in the regional jet streak is positioned near the upper-level divergent ¯ooding in the central plains. center in their composite so that the direct cross-jet cir- While the cloud-shield shape differs, both convective- culation enhances a portion of the upper-level diver- system types initiated in environments that are char- gence and, subsequently, accelerates a portion of the acterized by deep, synoptic-scale ascent and low-level low-level branch of the divergent circulation producing moisture transport from a southerly LLJ associated with a local wind maximum. This LLJ mechanism is con- continental-scale baroclinic waves. PECS occurrences sistent with the superposition of a divergent center as- initiated more often when vigorous waves exited the sociated with the continental-scale wave with an upper- intermountain region, whereas MCC occurrences initi- level jet streak in both MCC-relative and PECS-relative ated more often within a high-amplitude wave pattern composites. By positioning and enhancing the low-level with a trough positioned over the northwestern United moisture transport in this manner, the upper-level dy- States and a ridge positioned over the central Great namic forcing can indirectly in¯uence the position and Plains. development of MCC and PECS occurrences. Provided this upper-level support, an important ele- Notice that the gradient of the wind speed along the ment that organizes local convection into different types upper-lever jet streak axis is slightly less in the MCC- of convective systems (based on the shape of the cloud relative composite than in the PECS-relative composite. shield) appears to be the orientation of low-level thermal This means that either upper-level jet streaks are not as features. In the MCC-relative composite, convection is common, or they are not as strong. Despite this obser- focused to a particular region by low-level inversions vation, the southerly component in the MCC deviation south and west of the initiation position and an east±

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC 598 MONTHLY WEATHER REVIEW VOLUME 126

FIG. 15. Streamline analysis derived from the 850-mb deviation wind and lid-strength term (ЊC, shaded with interval of 1 starting with 1) for (a) MCC-relative and (b) PECS-relative composites.

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC MARCH 1998 ANDERSON AND ARRITT 599 west boundary north of the initiation position. In the 1989: A composite model of mesoscale convective complexes. PECS composite, the areal coverage of convection is Mon. Wea. Rev., 117, 765±783. Farrell, R. J., and T. N. Carlson, 1989: Evidence for the role of the not as effectively hindered by low-level inversions and lid and underrunning in an outbreak of tornadic thunderstorms. may be aided by a north±south thermal boundary. Mon. Wea. Rev., 117, 857±871. Fritsch, J. M., and R. A. Maddox, 1981: Convectively driven mesoscale pressure systems aloft. Part I: Observations. J. Climate Acknowledgments. This research was supported by Appl. Meteor., 20, 9±19. NSF Grants ATM-9627890 and ATM-9616728. We are , R. J. Kane, and C. R. Chelius, 1986: The contribution of me- grateful to John Augustine for providing the automated soscale convective weather systems to the warm-season precip- satellite classi®cation software and assisting with its im- itation of the United States. J. Climate Appl. Meteor., 25, 1333± 1345. plementation. We also thank Dr. Arlene Laing for useful Holton, J. R., 1992: An Introduction to Dynamic Meteorology. Ac- discussions on MCCs. Reviews by Dr. Laing and two ademic Press, 511 pp. anonymous reviewers considerably aided the presenta- Hoskins, B. J., I. Draghici, and H. C. Davies, 1978: A new look at tion and substance of this paper. This is Journal Paper the ␻-equation. Quart. J. Roy. Meteor. Soc., 104, 3±38. Kalnay E., M. Kanamitsu, and W. E. Baker, 1990: Global numerical J-17681 of the Iowa Agriculture and Home Economics weather prediction at the National Meteorological Center. Bull. Experiment Station, Ames, Iowa, Project No. 3245, and Amer. Meteor. Soc., 71, 1410±1428. supported by Hatch Act and State of Iowa funds. Kunkel, K. E., S. A. Changnon, and J. R. Angel, 1994: Climatic aspects of the 1993 upper Mississippi River basin ¯ood. Bull. Amer. Meteor. Soc., 75, 811±822. Laing, A. G., and J. M. Fritsch, 1997: The global population of REFERENCES mesoscale convective complexes. Quart. J. Roy. Meteor. Soc., 123, 389±405. Achtemeier, G. L., 1989: Modi®cation of a successive corrections Maddox, R. A., 1980: Mesoscale convective complexes. Bull. Amer. objective analysis for improved derivative calculations. Mon. Meteor. Soc., 61, 1374±1387. Wea. Rev., 117, 78±86. , 1983: Large-scale meteorological conditions associated with Arritt, R. W., T. D. Rink, M. Segal, D. P. Todey, C. A. Clark, M. J. midlatitude, mesoscale convective complexes. Mon. Wea. Rev., Mitchell, and K. M. Labas, 1997: The Great Plains low-level jet 111, 1475±1493. during the warm season of 1993. Mon. Wea. Rev., 125, 2176± , D. M. Rodgers, and K. W. Howard, 1982: Mesoscale convective 2192. complexes over the United States during 1981ÐAnnual sum- Augustine, J. A., 1985: An automated method for the documentation mary. Mon. Wea. Rev., 110, 1501±1514. of cloud-top characteristics of mesoscale convective systems. Menard, R. D., and J. M. Fritsch, 1989: A mesoscale convective NOAA Tech. Memo. ERL ESG-10, Dept. of Commerce, Boulder, complex-generated inertially stable warm core vortex. Mon. CO, 121 pp. [Available from NOAA/FLS, 325 Broadway, Boul- Wea. Rev., 117, 1237±1259. der, CO 80303.] Murray, R., and S. M. Daniels, 1953: Transverse ¯ow at entrance , and K. W. Howard, 1988: Mesoscale convective complexes and exit to jet streams. Quart. J. Roy. Meteor. Soc., 79, 236± over the United States during 1985. Mon. Wea. Rev., 116, 685± 241. 701. Rodgers, D. M., K. W. Howard, and E. C. Johnston, 1983: Mesoscale , and , 1991: Mesoscale convective complexes over the convective complexes over the United States during 1982. Mon. United States during 1986 and 1987. Mon. Wea. Rev., 119, 1575± Wea. Rev., 111, 2363±2369. 1589. , M. J. Magnano, and J. H. Arns, 1985: Mesoscale convective , and F. Caracena, 1994: Lower-tropospheric precursors to noc- complexes over the United States during 1983. Mon. Wea. Rev., turnal MCS development over the central United States. Wea. 113, 888±901. Forecasting, 9, 115±135. Schwartz, B. E., C. F. Chappell, W. E. Togstad, and X.-P. Zhong, Barnes, S. L., 1994: Applications of the Barnes objective analysis 1990: The Minneapolis ¯ash ¯ood: Meteorological analysis and operational response. Wea. Forecasting, 5, 3±21. scheme. Part III: Tuning for minimum error. J. Atmos. Oceanic Stensrud, D. J., 1996: Importance of low-level jets to climate: A Technol., 11, 1459±1479. review. J. Climate, 9, 1698±1711. Bartels, D. L., J. M. Skradski, and R. D. Menard, 1984: Mesoscale Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology. convective systems: A satellite-data-based climatology. NOAA Kluwer Academic, 660 pp. Tech. Memo. ERL ESG 8, Dept. of Commerce, Boulder, CO, Uccellini, L. W., 1980: On the role of upper tropospheric jet streaks 63 pp. [NTIS PB-85-187904.] and leeside cyclogenesis in the development of low-level jets in Bell, G. D., and J. E. Janowiak, 1995: Atmospheric circulation as- the Great Plains. Mon. Wea. Rev., 108, 1689±1696. sociated with the Midwest ¯oods of 1993. Bull. Amer. Meteor. , and D. R. Johnson, 1979: The coupling of upper and lower Soc., 76, 681±695. tropospheric jet streaks and implications for the development of Carlson, T. N., and R. J. Farrell, 1983: The lid strength index as an severe convective storms. Mon. Wea. Rev., 107, 682±703. aid in predicting severe local storms. Natl. Wea. Dig., 8, 769± Velasco, I., and J. M. Fritsch, 1987: Mesoscale convective complexes 774. in the Americas. J. Geophys. Res., 92, 9591±9613. Chen, T.-C., and J. A. Kpaeyah, 1993: The synoptic-scale environ- Zhang, D.-L., and J. M. Fritsch, 1987: Numerical simulation of the ment associated with the low-level jet of the Great Plains. Mon. meso-␤ scale structure and evolution of the 1977 Johnstown Wea. Rev., 121, 416±420. ¯ood. Part II: Inertially stable warm-core vortex and the me- Cotton, W. R., M.-S. Lin, R. L. McAnnelly, and C. J. Tremback, soscale convective complex. J. Atmos. Sci., 44, 2593±2611.

Unauthenticated | Downloaded 09/30/21 12:43 PM UTC